The Argaman-Kaufman flocculation model holds great promise as a practical tool for the generalization of flocculation behaviour from relatively few, simple batch tests in the laboratory. There are, however, valid reasons that have prevented it from being widely accepted in design practice. This paper deals with a number of these obstacles:
• There is the need of a standardized test method in terms of equipment, procedures and data analysis. The effects of settling time and the choice of the flocculation performance parameter are demonstrated.
• The flocculation constants are critically dependent on the level of coagulant dosage. In the sweep flocculation regime, the optimum coagulant dosage cannot be precisely pinpointed with jar testing, which presents a formidable problem for routine application.
• The flocculation constants are, contrary to the model, dependent on the initial particle concentration. This currently precludes the extrapolation of test results to raw water turbidities different from the turbidity at which the test was conducted.
• The flocculation performance parameter is not an intuitive parameter in design practice, and the design procedure for flocculation systems will have to be adapted to reach a specific design objective.